May 2003

Executive Summary

Key words: Transitional Water, Coastal Water, Typology, Water Framework Directive.

The purpose of this contract was to deliver a common typology for the UK and Ireland's transitional and coastal waters, funded jointly through SNIFFER (Scotland and N. Ireland Forum for Environmental Research), the Environment Agency of England and Wales and the Irish Environmental Protection Agency. The work was completed by a consortium which consisted of Aqua-Fact International Services Ltd, British Geological Survey, the Centre for Environment, Fisheries and Aquaculture Science, the Institute of Estuarine and Coastal Studies, and the Joint Nature Conservation Committee.

For England, Wales, Scotland and Ireland, the coastal waters out to 3 nautical miles, and the transitional waters as defined by the Urban Waste Water Treatment Directive, were gridded using the GIS software ArcView. For the first phase of data analysis there were 30,720 cells in coastal waters, and 196,728 cells in 179 transitional waters. The proportions of each category of each physical factor in a cell or group of cells were clustered and analysed using Multiple Correspondence Analysis. Repeat analyses were undertaken using improved data for some physical factors, and using revised boundaries for some transitional waters.

The final physical typology of transitional waters described the majority of estuaries as a continuum within which there were two major groups. One of these groups represented estuaries that were generally partly mixed or stratified, and which had a tendency to be mesohaline or polyhaline. The other major group of estuaries were fully mixed, polyhaline or euhaline estuaries which were sheltered, generally had a sand or mud substratum and tended to have extensive intertidal areas. The main conclusion of this typology process was that the multivariate techniques allocated most transitional waters to a continuum that described typical estuarine features with extremes at each end, and with a number of small satellite groups with very specific characteristics. The biological validation did not identify significant differences in the biotope complex composition of the two main physical types, but it was possible to characterise and describe these two main types. An alternative biological typology was developed using a comprehensive biological dataset and expert judgment that identified five distinct biological types of transitional waters.

Populating 0.1 nm cells in transitional waters involved a large amount of interpolation and extrapolation. For example, salinity was often not known in detail and major assumptions were made to populate cells. Similarly, mixing and residence time was calculated mathematically from knowledge of the freshwater inputs and size of the estuary. Given that many of the variables were estimated, averaged across cells and block-filled based on limited data, it is not surprising that many estuaries were relatively similar.

The results of the initial coastal water analysis followed a similar pattern, however, enlarging the coastal cell size to 10x10nm and reducing and dichotomising the physical factors simplified the analysis and resulted in an acceptable typology. The analysis generated three dominant groups, all of which were characterised by being euhaline, but which had different characteristics of wave exposure and current velocity. One group was widespread in the coastal and offshore waters of W Scotland, W and SW Ireland, and SW England, and described euhaline regions of generally high wave exposure and low current velocity. Another described areas in E and SE Ireland, NE and SE England, which were euhaline and generally characterised by low wave exposure and low current velocity. The third major group described euhaline waters that had a moderate or strong current velocity and high levels of wave exposure. Evidence from a range of biological sampling programmes suggested that this coastal typology could be broadly supported.

One of the major problems with this typology programme has been the need to identify types using physical data, while independently validating these with biological data. One technique which could have generated types of estuaries and coasts that more closely agreed with preconceptions would be the use of multivariate analysis on both physical and biological data at the same time, so that groups of similar areas could be selected and explained using both sets of data. Allowing both data sets to have an equal contribution to the final typology would provide a method for biological validation that was an integral part of the typology process.

The relative lack of success with biological validation of whole estuaries suggests that the process should be extended to test the types at the finer community level. It may even be necessary to generate types for each sub-area of an estuary. Such an approach should reveal whether the ecological character of a type, at the level at which reference conditions and classification in quality are likely to be addressed, can confirm the ecological validity of the final typology.

N.B. The report is available for download from the SNIFFER Website